The histopathology of cancer suggests a progressive loss of differentiation, where mature cells are replaced by less-differentiated counterparts with embryonic behaviors such as high rates of growth and cell migration. Some of the genes controlling these functions may act in a tissue-specific manner and potentially function as tumor suppressor genes. It is further hypothesized that new tissue-specific tumor suppressors may be found by the histopathologic examination of mutant embryos of the zebrafish (Danio rerio), for abnormal differentiation in specific, mature embryonic tissues. The experimental features of the zebrafish make it a unique vertebrate model to find mutants defective in differentiation of specific tissues. First, each female is capable of producing an average of about 100 rapidly developing, 1 x 1 x 3 mm, transparent embryos per week, making it possible to readily produce and examine thousands of mutant candidates ex vivo, at a relatively low cost. Parthenogenesis can be used to unmask recessive mutations within one generation in heterozygous parents. Lethal mutations can also be studied, since the heterozygous parents allow one to generate additional mutants for further study or genetic mapping. Recent progress in zebrafish genetics makes positional cloning possible. The production of many early embryonic developmental mutants by other investigators indicates that the zebrafish is a practical vertebrate model for generating interesting genetic mutations. These studies are expected to reveal genes important to differentiation of specific tissues in the zebrafish. Since human cancers show abnormal differentiation, homologues of some of these genes may be affected in human cancer. It is hoped that work on such genes, their products, and expression, will lead to prognostic and therapeutic tools for patient care.